From Prediction to Governed Intervention: AI-Enabled Construction Project Controls for Productivity, Resilience and Net-Zero-Oriented Delivery

Artificial intelligence (AI) is increasingly used in construction to forecast duration, monitor progress, prioritise risk, support procurement and logistics, improve supply chain visibility, and compare environmental trade-offs. These applications are often judged by their technical performance, yet that does not show whether, or how, an analytical output changes a project decision. This paper addresses that gap through governed decision translation: the process by which an AI-enabled output is inter-preted, validated, challenged, authorized, assigned for implementation, documented, and reviewed. A structured integrative review with framework synthesis was con-ducted using a ScienceDirect seed stream and targeted Web of Science cross-checks. The 34-study corpus was classified by evidence relevance and appraised across study design, deployment maturity, outcome proximity, and methodological credibility. Most studies focus on forecasting, monitoring, optimization, and decision support. Only one provides clear evidence of implementation in a live project, while two others approach an identifiable project-control intervention pathway. The evidence therefore supports theory building rather than causal claims of performance improvement. The resulting Governed AI Project Controls Framework distinguishes the data and analytical foundations of AI-enabled control from the decision interface, governed translation, authorised intervention, value domains, and subsequent learning. It also separates structural, procedural, and interpretive governance. Productivity related performance is treated as relatively close to intervention, resilience as a dynamic capability, and net-zero-oriented delivery as a more cumulative environmental domain. The evidence supports carbon-, energy-, waste-, and material-aware decisions, but not claims of achieved net-zero delivery. The paper offers a conditional explanation of how AI-supported insights may acquire authority, operational consequence, and accountability in temporary, multi-organisational construction projects, together with propositions, boundary conditions, and observable indicators for empirical testing.